Optimal sequences for non‐phosphate‐directed phoshorylation by protein kinase CK1 (casein kinase‐1) – a re‐evaluation

A variety of synthetic peptides derived from either the inhibitor‐2 (I‐2) phosphoacceptor sites or the optimal sequences selected in an oriented peptide library have been compared for their susceptibility to phosphorylation by protein kinase CK1 (also termed casein kinase‐1). The I‐2‐derived peptides are by far preferred over the library peptides by both rat liver CK1 (and by the α/β, γ and δ/ε isoforms immunoprecipitated from it) and recombinant Xenopus laevis CK1α. The superiority of the I‐2‐derived peptides over the library ones is reflected by V max values one to two orders of magnitude higher while the K m values are comparable. Individual substitutions of any of the aspartic acids with alanine in the I‐2‐derived peptide RRKHAAIGDDDDAY S ITA is detrimental, producing both a fall in V max and an increase in K m which are more pronounced at position n  –3, but also quite significant at positions n  –4, n  –5 and, to a lesser extent, n  –6. The unfavourable effect of these substitutions is more evident with rat liver CK1 than with recombinant Xenopus laevis CK1α. The chimeric peptide IGDDDDAY‐ S‐ IIIFFA , resulting from the combination of the N‐terminal acidic sequence of the I‐2 (Ser86) site and the C‐terminal hydrophobic cluster selected in the library peptides (MAEFDTG‐S‐ IIIFFA KKK and MAYYDAA‐S‐ IIIFFA KKK) is phosphorylated as efficiently as the I‐2‐derived peptide in terms of both K m and V max . These combined data strongly support the conclusion that, at variance with the optimal sequences selected in the library, optimal non‐phosphate‐directed phosphorylation of peptide substrates by CK1 critically relies on the presence of a cluster of acidic residues (preferably aspartic acid) upstream from position n – 2, while the highly hydrophobic region downstream from serine selected in the library appears to be dispensable. The reason for these discrepancies remains unclear. The possibility that the library data are biased by the invariant elements forming its scaffold (MA‐x‐x‐x‐x‐x‐SI‐x‐x‐x‐x‐AKKK) would be consistent with the observation that the library‐selected peptides, despite their low K m values, fail to compete against the phosphorylation of protein and peptide substrates by CK1, suggesting that they bind to elements partially distinct from those responsible for substrate recognition.